The transition behaviour from school to studies – Empirical evidence on the reasons of a delayed start of studies in Germany

With the end of school and the diploma qualifying for university admission or matriculation young people in Germany can choose between vocational education and studies. In the period from 2000 to 2015 the share in young people that decided in favour of studies increased from 71% to 74% implying that the number of first-year students grew in average by 2.8% p.a. However, the decision in favour for a study does not necessarily coincide with the actual start of the studies. The data also shows that in 2011 62% of the young people who chose to study started their studies right after school. 24% waited one year and 14% waited two years or longer. Compared to the past there seems to be a tendency towards shortening the time lag, i.e. a growing number of people starts the studies right away. The supply of university places hence has to be adapted to changes in the demand and depends on the propensity to study and the preferred starting point of the studies. While the reasons for the decision in favour of studies are rather well known and documented, the reasons for time lags in the start of studies were less subject to empirical investigation. Moreover, modelling the transition from end of school to the first year of studies based on empirical findings can improve to calculate the required number of university places. The model for estimating the expected number of first-year students and needed university places was developed by the Standing Conference of Education Ministers (KMK 2014) and is based on many status quo assumptions. Regarding the transition time between completion of school and start of studies constant values were assumed in the base model. The transition behaviour stays constant throughout the projection period (2016-2025) and equals the average value of the last three years (2013-2015) of the historical data set (1992-2015). Conducting a sensitivity analysis helps to understand the impact of this constancy assumption for the model results. Using econometric analysis the constancy assumption then can be relaxed. In the alternative model the time lag will endogenously depend on time trends or on other explanatory variables. Different methods (time series analysis with time trends, Probit regression model, Panel data analysis) will be applied, tested and compared. The best alternative approach will be used to compare it to the base model. The conservative approach in assuming a constant transition (base model) leads to a considerably lower supply of university places than the alternative model considering behavioural changes. The results suggest a gap between underlying demand and actual supply. As a consequence, a lot of students cannot start studying because of the mismatch and postpone the start of studies, by this partly confirming the conservative expectations. By allowing the transition to be endogenously dependent on time trends and other variables the original projection model can be improved. Comparing the base model with the alternative model differences in the number of first-year students and the related university places can be quantified. Overall, this offers the opportunity to identify deficiencies and to adjust the planning.